R Notebook

###change

library(leaflet)
library(readxl)
library(dplyr)
library(reticulate)

maindata <- read_excel("maindata.xlsx")
tmp=maindata

Let look at how many unique values each column has

sapply(maindata, function(x) length(unique(x)))

                    id              log_price          property_type              room_type 
                 74111                    767                     35                      3 
             amenities           accommodates              bathrooms               bed_type 
                 67122                     16                     18                      5 
   cancellation_policy           cleaning_fee                   city            Description 
                     5                      2                      6                  73469 
          first_review   host_has_profile_pic host_identity_verified     host_response_rate 
                  2555                      3                      3                     81 
            host_since       instant_bookable            last_review                    lat 
                  3088                      2                   1372                  74058 
                  long                   name          neighbourhood      number_of_reviews 
                 73973                  73331                    620                    371 
  review_scores_rating          thumbnail_url                zipcode               bedrooms 
                    55                  65884                    669                     12 
                  beds 
                    19 

maindata %>% group_by(cancellation_policy) %>% select(cancellation_policy)%>% unique()

count(maindata, cancellation_policy) %>% mutate(relative_freq=(n/sum(n)))

maindata %>% group_by(bed_type) %>% select(bed_type)%>% unique()

library(dplyr)
maindata %>% select(property_type) %>% count(property_type) %>% mutate(relative_freq=(n/sum(n)))

maindata %>% filter(city=="NYC") %>% select(bed_type) %>% count(bed_type) %>% mutate(relative_freq=(n/sum(n)))

library(ggplot2)
maindata%>% ggplot( aes(x=cancellation_policy, y=log_price))+ geom_boxplot()+ ggtitle( "Boxplot for log_price vs cancellation policy")

ggplot(maindata, aes(x=bed_type, y=log_price))+ geom_boxplot()+ ggtitle( "Boxplot for log_price vs bed type")

ggplot(maindata, aes(x=cleaning_fee, y=log_price))+ geom_boxplot()+ ggtitle( "Boxplot for log_price vs cleaning fee")

Do some hypothesis testing.

vcd::mosaic(~cleaning_fee+instant_bookable, data=maindata, shade=TRUE)

First, let us plot the different properties, using the latitude and longitude information given in our dataset.

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat, color = "red",
stroke=FALSE)
m

From the map above, we can see that the properties listed in our datset are from six different locations: Los Angeles, New York, DC, Boston, Chicago, San Fransico etc.

log_price for th different cities.

NYC

rows = (maindata$city== "NYC") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$log_price,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat,radius = ~log_price,
fillColor = ~mypalette(log_price),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~log_price, opacity=0.9,
title = "Log_price", position = "bottomright" )
m

Talk about which part of NYC has higher log_price and which part has lowest.

Boston.

rows = (maindata$city== "Boston") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$log_price,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat,radius = ~log_price,
fillColor = ~mypalette(log_price),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~log_price, opacity=0.9,
title = "Log_price", position = "bottomright" )
m

DC

rows = (maindata$city== "DC") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$log_price,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat,radius = ~log_price,
fillColor = ~mypalette(log_price),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~log_price, opacity=0.9,
title = "Log_price", position = "bottomright" )
m

San Francisco.

rows = (maindata$city== "SF") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$log_price,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat,radius = ~log_price,
fillColor = ~mypalette(log_price),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~log_price, opacity=0.9,
title = "Log_price", position = "bottomright" )
m

LA

rows = (maindata$city== "LA") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$log_price,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat,radius = ~log_price,
fillColor = ~mypalette(log_price),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~log_price, opacity=0.9,
title = "Log_price", position = "bottomright" )
m

We can see that along the edge there are properties that are high priced.

Chicago

rows = (maindata$city== "Chicago") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$log_price,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~long, ~lat,radius = ~log_price,
fillColor = ~mypalette(log_price),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~log_price, opacity=0.9,
title = "Log_price", position = "bottomright" )
m

Types of property

rows = (maindata$city== "Chicago") 
tmp = maindata[rows, ]

mybins <- seq(2, 8, by=1.5)
mypalette <- colorBin( palette="YlOrBr",
domain=tmp$property_type,
na.color="transparent",
bins=mybins)

m <- leaflet(tmp) %>%
addTiles() %>%
addProviderTiles("OpenStreetMap.BZH") %>%
addCircleMarkers(~longitute, ~latitude,radius = ~log_price,
fillColor = ~mypalette(property_type),
fillOpacity = 0.5,
color = "white",stroke=FALSE
)%>%
addLegend( pal=mypalette, values=~property_type, opacity=0.9,
title = "Property type", position = "bottomright" )

Error in eval(f[[2]], metaData(data), environment(f)) : 
  object 'longitute' not found